Hi everyone, I have a question and

[LUAY_89]

Hi everyone, I have a question and need your help. We had plate with actual thickness of 0.75” Rolled to 84” ID shell. I checked the circumference and ID of shell after rolling and fit-up long seam was 268 7/8” and 84 3/32”ID respectively.The circumference required to get 84”ID must be 268 5/8” but the actual circumference I have is 1/4” bigger. My question is do we have tolerance in ASME Vlll Div.1 that can be applied on circumference of rolled shell? The rolled shell should meet UG-80 and UG-78.

Thanks!

 

[DriveMeNuts]

You should look at the fabrication drawing for the allowable tolerance. If the tolerance has not been defined before the start of fabrication, then in my opinion this is a sign of incompetence.

I think ASME VIII Div 1 U-5 will answer your question.

 

[crshears]

Hello LUAY 89 et al,

As a matter of etiquette, may I politely request that when posting to these fora you please provide a subject for ready reference, and do not leave the subject line blank, as in these cases the first few words of your post become the default subject, and it can lead to difficulties for those wishing to respond to differentiate between all of the various posts on numerous topics that all nevertheless begin with the subject "Hi everyone, I have a question . . . "

If you won't police yourselves in this, the eng-tips overlords may be forced to modify the site's software to disallow posts with a blank subject line, and I for one do not wish to challenge the unknown limits of their beneficence.

 

[r6155]

plain plate before rolling is (ID + plate thickness) x PI – bevel gap

(84 + 0.75) x PI – 0.0625 = 266.1875”

Bevel gap depends of your weld design.

Regards
r6155

 

[SnTMan]

LUAY_89. Sec VIII, Div 1 has rules for tolerances on roundness of the cylinder, UG-80 as you noted, but no rules as to tolerances on the actual diameter. See U-5.

Tolerances not addressed by the Code are normally specified in industry standards or client specs. Actually 3/32 on 84 inch if welded out is not bad. Often, bigger is better than smaller, as whatever is supposed to fit inside will fit inside.

BTW your ref to UG-78 appears wrong to me.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

Tolerance is not the problem here. Wrong dimension of plate before rolling is the problem, neutral axis of plate was not taken into account.

Regards
r6155

 

[david339933]

r6155, Although I have the same hunch, what is telling you it is the wrong cut length of plate? There is no cut length listed, only OD circ.

 

[SnTMan]

The question was about tolerance.

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

The intention (design) of LUAY is rolling an ID 84" shell with 0.75" plate, then the required cut length is 266.1875".

Different cut length will result in mistake.

Tolerance shall be applied on the design dimension, not on the mistake.

Regards
r6155

 

[SnTMan]

So, it is within tolerance or no?

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

LUAY accounts are inconsistent.

If actual ID of shell is bigger than the design ID (84”) check the required thickness. If less than 0.75” then is OK.

ID of shell is equal to ID of heads? See UW-33

Good engineering practice is to measure actual ID of heads and after this cut the length of shell plate.

Regards
r6155

 

[SnTMan]

So, it is within tolerance or no?

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

Due that tolerance of diameter is not specified in ASME VIII then is not applied here.
My engineering judgment is clear in my comments, like as mentioned in U-5

Regards
r6155

 

[SnTMan]

Agreed. If the work were to TEMA the tolerance would be 1/8" plus tolerance only. So within. If.



The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[LUAY_89]

The length of plate after cutting was 266 1/4" with 1/16" out of squareness and due to stretching after rolling, the shell circumference is 1/4" bigger than required. As per UG-80 The Max ID - Min ID should be within 1% which is 0.84" and the max & Min ID I got are 84 3/4" and 83 5/16" respectively. As a result, the difference is 1.74" which is out of UG-80 tolerance.

All I need now is someone can confirm my calculation above.
Thank you for your prompt response.

 

[LUAY_89]

Just one more question Can we apply UG-80 on Half shells or only on completed shells? If not Does ASME VIII Div. 1 have a tolerance on half shells?

 

[SnTMan]

LUAY_89 , 0.01 * 84 = 0.84 confirmed, 84 12/16 - 83 5/16 = 1 7/16, 1.74 not confirmed, UG-80 limit out-of round exceeded, confirmed.

I doubt UG-80 or any Code rule applies to half cylinders. How would out-of-round be measured?

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

SnTMan. more information.
a)The TEMA Mechanical Standards are applicable to shell and tube heat exchangers whichdo not exceed any of the following criteria:
(1) inside diameters of 100 in. (2540 mm)

b)PD 5500
Circumferential tolerance
Up to and including 650 mm OD ±5 mm
Over 650 mm OD ±0.25% of circumference

c)BS-EN 13445
5.4 Tolerances for vessels subjected to internal pressure
5.4.1 External diameter
For cylindrical and spherical pressure vessels the mean external diameter derived from the circumference shall not
deviate by more than 1,5 % from the specified external diameter.

LUAY_89. I have doubts about your measurements. 1.74" is in the same section?, see UG-80.

You can apply UG-80 on half shells, measurements shall be made from a segmental circular template. See UG-80(b).

Regards
r6155

 

[DriveMeNuts]

The diameter deviation is negligible.

The out of roundness can surely be corrected with clamps when welding the cylinder to its head or flange.
Or perhaps place the cylinder back in the rolling machine, to improve the ovality.

 

[r6155]

Rolling machine is the unique solution.
I suggest RT 10% of the longitudinal weld before rolling and RT 100% after rolling. This weld shall be flush with base metal and free of undercuts before RT.

Regards
r6155